STMicroelectronics N.V. (STM) Q3 2026 Earnings Report, Transcript and Summary

Ladies and gentlemen, welcome to the STMicroelectronics Third Quarter 2026 Earnings Release Conference Call and Live Webcast. I am Sandra, the Chorus Call operator. [Operator Instructions] The conference is being recorded. The conference must not be recorded for publication or broadcast. At this time, it is my pleasure to hand over to Jerome Ramel, EVP, Corporate Development and Integrated External Communications. Please go ahead, sir.

Thanks, Sandra, and thank you, everyone, for joining our third quarter 2026 financial results call. Hosting the call today is Jean-Marc Chery, ST President and Chief Executive Officer; joining Jean-Marc on the call are Lorenzo Grandi, President and CFO; and Remi El-Ouazzane, President, Microcontroller, Digital ICs and RF Product Group. This live webcast and presentation materials can be accessed on ST Investor Relations website. A replay will be available shortly after the conclusion of this call. This call will include forward-looking statements that involve risk factors that could cause ST results to differ materially from management expectations and plans. We encourage you to review the safe harbor statement contained in the press release that was issued with the results this morning and also in ST's most recent regulatory filings for a full description of these risk factors. [Operator Instructions] Now I'd like to turn the call over to Jean-Marc Chery, ST President and CEO.

Thank you, Jerome. Good morning, everyone, and thank you for joining ST for our Q3 2026 earnings conference call. I will start with an overview of the third quarter, including business dynamics. And I will hand over to Lorenzo for the detailed financial overview. I will then comment on the outlook and conclude before answering your question. So starting with Q3. Our third quarter net revenues of $3.76 billion came in above the midpoint of our business outlook range, mainly due to higher revenue in Communication Equipment and Computer Peripheral, driven by AI datacenter and lower Earth orbit satellite revenues, communication and personal electronics. Gross margin was 37.3%, above the midpoint of our business outlook range. Non-U.S. GAAP diluted earnings per share was $0.37. During the third quarter, inventory in our balance sheet increased slightly to support the expected increase in Q4 revenues. In distribution, inventory decreased and continued to track below our standard target. We generated positive free cash flow of $351 million. Let's now discuss our business dynamics during Q3. During the quarter, the ongoing recovery and secular growth across the end markets we serve continued to gain momentum. Demand strengthened further in Q3 with bookings increasing sequentially and book-to-bill remaining well above 2 overall. Book-to-bill was well above 1 in all end markets and significantly above 2 in Communication Equipment and Computer Peripheral and Industrial. Visibility improved. Inventory distribution continued to track below our standard target and demand continued to outpace supply across multiple product categories. Automotive revenues were stable sequentially and increased 5% year-over-year. Our solid position in automotive continued to be driven by application-specific ICs and sensors for core application, electrical powertrain and ADAS. We maintained our automotive design win momentum with OEMs and Tier 1 suppliers across the portfolio of ST product families, including microcontrollers and Power & Discrete. These wins covered hybrid, electric and internal combustion engine vehicles in areas including onboard charging, powertrain and active suspension as well as a first award for 48-volt electromechanical braking. We made progress with our automotive MEMS business, supported by the successful integration of NXP's MEMS sensors, achieving several key customer awards in telematics, cockpit control and tire pressure monitoring. We had another important design win with our in-cabin monitoring solutions, and we introduced a new automotive image sensor for infrared in-cabin sensing, designed to extend driver and occupant monitoring across more vehicle models. Industrial revenues improved 12% sequentially and 40% year-over-year. This solid growth was mainly driven by our general purpose microcontrollers, analog and power conversion products. During Q3, the broad-based recovery in industrial continued and inventory in distribution declined further remaining below our standard target. We had design wins across industrial automation, robotics, power systems, building automation, power tools and home appliances. Our portfolio combines intelligent sensing, real-time control and efficient power management to support the transformation of factory automation, robotics and power energy infrastructures and the growing physical AI market. At the Embedded World North America event, we had a demonstration together with NVIDIA, illustrating our broad cooperation on robotics. We showed an integrated robotics development path from digital twin to deployed application based on our inertial sensors, microcontrollers and imaging solutions. We are continuing to strengthen these capabilities with new products. Our latest launch is a miniature 6-axis motion sensor with a machine learning core designed to improve accident alerts, robot teaching, board stabilization and structural monitoring. By bringing AI processing into the sensor, systems can reduce energy use and save resources. In embedded processing, we continue to expand our STM32 offering with a strong pipeline of new products and software innovations. You will hear more about this at our STM32 Summit on November 3, followed by a dedicated STM32 analyst call on November 18. For Personal Electronics, third quarter revenues came in better than expected, up 12% sequentially and down 4% year-over-year. During the quarter, we had design wins with secure solutions, MEMS and imaging sensors and microcontrollers across a variety of devices, including smartphones, gaming headsets, earbuds and smart rings. For Communication Equipment and Computer Peripherals, third quarter revenues were above expectation, increasing 14% sequentially and 64% year-over-year. This strong growth mainly reflected the contribution of our company-specific growth drivers in AI datacenter and low Earth orbit satellite communications. For AI datacenters, we continue to see strong demand driven mainly by our silicon photonic ICs, electronic ICs and microcontrollers as well as power solutions. This supports our revenue expectation for 2026 of about $1.3 billion and assuming the current dynamic continues and with the current engagements we have, our revenue ambition of well above $2 billion in 2027 and further growth beyond. Now over to Lorenzo, who will present our key financial figures.

Thank you, Jean-Marc. Good morning, everyone. Let's start with a detailed review of the third quarter, starting with revenues on a year-over-year basis by reportable segment. Analog products, MEMS and Sensor grew 8.8%, mainly due to MEMS and to a lesser extent, analog and imaging. Power & Discrete products increased 11.3%. Embedded Processing revenues were up 24.1%, mainly due to general purpose microcontrollers and RF & Optical Communication grew 46.2% -- by end market, Communication Equipment and Computer Peripheral grew 64%; Industrial, 40%; and Automotive 5%, while Personal Electronics decreased 4%. Year-over-year, sales to OEMs and distribution increased 10.2% and 38.8%, respectively. On a sequential basis, by reportable segment, Analog Product, MEMS and Sensor increased by 9.4%, Power & Discrete by 2.8%, Embedded Processing by 5.7% and RF Optical Communication by 13.4% -- by end market on a sequential basis, Communication Equipment and Computer and Peripheral grew 14%, Industrial, 12% Personal Electronic, 12%, while Automotive was flat. Turning now to profitability. Gross profit in the third quarter was $1.40 billion, increasing 32.2% on a year-over-year basis. Gross margin was 37.3%, increasing 410 basis points year-over-year, mainly due to a lower unused capacity charges and better product mix. On a sequential basis, gross margin increased by 250 basis points. Q3 gross margin included about 60 basis points of unused capacity charges and about 40 basis points of negative impact resulting from nonrecurring cost related to our manufacturing reshaping program. Looking beyond these near-term effects, improving our gross margin remains a key priority. We remain committed to achieving the 44% to the 46% gross margin of our intermediate financial model once our manufacturing reshaping plan is completed. Starting from the approximately 39% gross margin expected for Q4 '26, we see a clear path to margin expansion. First, Q4 '26 temporary headwinds, including approximately 40 basis points from unused capacity charges and approximately 30 basis points from nonrecurring costs associated with our manufacturing reshaping program are expected to largely disappear during the second half of 2027. Second, we continue to expect more than 400 basis points of structural gross margin improvement from the reshaping of our manufacturing footprint. This benefit is expected to materialize as a step function either toward the end of 2027 or in early 2028, subject to the shutdown of legacy fabs and customers' qualification timing. Finally, we expect better mix and price to contribute to margin expansion over time, while input cost increase is expected to be a headwind. The timing and the magnitude of these effects remain more difficult to assess. Moving now to total net operating expenses. In the third quarter, this amounted to $979 million, excluding restructuring. Non-U.S. GAAP OpEx stood at $972 million, better than the indication we gave in July, reflecting continuing cost discipline. For the fourth quarter of 2026, we expect non-U.S. GAAP net OpEx to stand at about $1.035 billion. The sequential increase is mainly due to the calendar days effect net of vacation. This would lead the full year 2026 non-U.S. GAAP net OpEx slightly above $3.8 billion, in line with our expectation. In the third quarter, we reported operating income of $367 million, which includes $55 million for impairment, restructuring charges and other related phaseout costs. These charges are related to the execution of our company-wide program to reshape our manufacturing footprint and to resize our global cost base. Q3 operating income also included $7 million in purchase price allocation effects from our acquisition of NXP's MEMS sensor business. Excluding these items, Q3 non-U.S. GAAP operating income stood at $429 million and non-U.S. GAAP operating margin was 11.4%, with Analog Products, MEMS and Sensor at 14.1% Power & Discrete negative 17.1% Embedded Processing at 23.1% and RF Optical Communications at 26.2%. Third quarter 2026 net income was $291 million compared to $237 million in the year-ago quarter. Diluted earnings per share were $0.31 compared to $0.26. Non-U.S. GAAP net income stood at $344 million and non-U.S. GAAP diluted earnings per share stood at $0.37. Net cash from operating activities totaled $1.02 billion in the third quarter compared to the $549 million in the year-ago quarter. Net CapEx was $643 million in the third quarter compared to $401 million in the year-ago quarter. Free cash flow was positive at $351 million in the third quarter compared to the positive $130 million in Q3 2025. Inventory at the end of the quarter was $3.25 billion compared to $3.19 billion in Q2 2026 and $3.17 billion in Q3 2025. Days sales of inventory at quarter end were 126 days, better than our expectation, unchanged from the previous quarter and compared to the 135 days in the year-ago quarter. Cash dividends paid to stockholders in the third quarter of 2026 totaled $83 million. In addition, ST executed a share buyback of $91 million. ST maintained its financial strength with a net financial position that remains solid at $2.11 billion as of September 26, 2026, reflecting total liquidity of $5.31 billion and total financial debt of $3.21 billion. Following the exercise of the call option for the early redemption of ST's convertible debt due in 2027, bondholders exercised their conversion rights on the total $750 million convertible bonds. ST settled the bonds on a net share basis, delivering $750 million in cash and about 4.2 million share from treasury stock. Finally, while we are not providing a quarterly indication for 2027, for modeling purposes, please note that 2027 will have 86 calendar days in Q1, 91 in Q2, 91 in Q3 and 97 in Q4. Now back to Jean-Marc, who will comment on our outlook.

Thank you, Lorenzo. Now let's move to our business outlook for Q4 2026. We are expecting Q4 revenues of $4.15 billion, plus/minus 350 basis points at the midpoint. Our Q4 '26 net revenues will increase 10.5% sequentially and 24.7% year-over-year. We expect our gross margin to be at about 39%, plus/minus 200 basis points, including about 40 basis points of unused capacity charges. This business outlook does not include any impact of potential further changes to global trade tariffs compared to the current situation. For 2026, the midpoint of this outlook translates into full year 2026 revenues of about $14.5 billion, representing a 22.8% year-over-year increase. Beyond the cyclical recovery across our end markets, the year-over-year growth is supported by ST's company-specific growth drivers, including AI datacenters and low Earth orbit satellite communications. Gross margin is expected to be about 36.4%. Non-U.S. GAAP gross margin is expected to be about 36.6%, representing an improvement of 270 basis points versus the fiscal year 2025. To conclude, we continue to see a very strong demand for our AI datacenters and low Earth orbit satellite communications products and technologies, alongside an ongoing cyclical recovery in automotive and Industrial. With demand across many of the markets we serve exceeding available supply, we are expanding capacity in line with ST's strategic road map and growth opportunities. At the same time, improving profitability remains a key priority as we progress toward our intermediate financial model. As mentioned earlier by Lorenzo, this is supported by our company-specific growth drivers, combined with the execution of our company-wide program to reshape our manufacturing footprint and resize our global cost base. Before handing over to Jerome, I would like to mention a few corporate developments. Earlier this month, the Supervisory Board asked me to be reappointed in my current role for an additional year. I was honored and pleased to accept the proposal. My reappointment as Chairman of the Managing Board and ST's President and CEO will be proposed for shareholder approval at ST's 2027 Annual General Meeting of Shareholders. Finally, earlier today, we announced updates to ST's organization aimed at better aligning our vision and business objectives with evolving market dynamics. This will strengthen the coordination between our strategy, technology and product road maps, accelerate innovation and time to market, ultimately supporting our long-term growth opportunities. Remi El-Ouazzane is confirmed as President, member of the Executive Committee and Head of Microcontrollers, Digital ICs, RF and Optical Communication products. He will also take on leadership on the Imaging Sensor Group and be appointed as the Chief Technology Officer. Marco Cassis is confirmed as President, Member of the Executive Committee and Head of Analog, Power & Discrete and MEMS Product. He will also take on leadership of R&D for Analog, Power & Discrete and MEMS technologies and be appointed as the Chief Strategy Officer. The new organization and the related adjustment to our reportable segments will take effect on January 1, 2027 with implementation of the changes to be completed by the end of March 2027. Thank you, and we are now ready to answer your questions.

[Operator Instructions] Our first question comes from Joshua Buchalter from TD Cowen.

On the results. I guess to start, I want to ask about some of the comments you made on gross margins. So I believe you said after the manufacturing transition headwinds abate, there'd be a step function increase of 400 basis points to gross margin. Is that sort of going to hit all at once? Because I think you previously said we shouldn't expect gross margins to get to 40% until sort of second half '27 or first half '28. So I guess is the expectation that it very quickly goes from 40% to 44%?

Maybe I take this question. Clearly, in terms of gross margin, we -- what we said is that the impact of our reshaping plan in manufacturing, it will factor in, in our gross margin once we will close our fabs. You know that there are 4 fabs, the legacy fabs that we intend, let's say, to close. And this, let's say, clearly, the impact, the positive impact will be visible when we will, let's say, be able to close. And this will happen in the second part of next year, let's say, so means that it will happen towards the end of 2027 and let's say, at the beginning of 2028. Clearly, here, it depends on the acceptance of the customer for 3 of these fabs substantially, let's say, we believe we have all in our hands, let's say, still some discussion for one of those. In this respect, I would say that in any case, we will continue to see improvement of our gross margin moving forward along the various quarter also due to the fact that we have a positive impact of the mix and improvement in terms of manufacturing variance. As I was saying, progressively, let's say, reducing our unloading charges and reducing also the temporary impact of negative cost due to the activity to transfer product from the legacy fab to the new fab 300-millimeter or in case of silicon carbide to the 200 millimeter.

Okay. My follow-up, any help you can give us with either like context or any sort of tighter range on the datacenter contribution for next year of well greater than $2 billion? And any help you can give us on how much you expect to be MCUs versus electronic ICs and PICs and power?

On the revenues, I assume.

Josh, your question was on revenues?

Yes. Revenue contribution from datacenter.

Okay. So I will give the overall picture and Remi will comment on the detail. Well, as I said during my speech, this year, we will generate about $1.3 billion revenue for datacenter, 85% coming from MDRF, so mainly microcontroller, silicon photonics and electronic ICs and BiCMOS. Next year, I will not give in detail, okay? We will be well above doubling this number. And we will communicate in January, of course, a number for '27 and an indication for '28. So maybe, okay, Remi, you can comment some more detail.

Sure. Microcontrollers in datacenters will see a steep growth year-on-year. It will grow 2x in revenue size. I won't give you the exact number. The contribution of silicon photonics next year will start to grow and become very relevant. I would say that actually, it will be tens of percent of the complete revenue consideration. We will not share the exact details. As to BiCMOS, I think the bulk of the effect we will see is from '28 onwards because we've won a lot of sockets over the past quarter. And I think that you'll see the effects mainly in '28, even though it will be nonnegligible in '27, but I guess from a contribution standpoint, it will be very, very visible in '28.

Thank you, Remi. So about $1.3 billion this year, well above doubling next year, and we continue towards '28, and we will communicate more detail in January.

The next question comes from Francois Bouvignies from UBS.

I had just a question on the gross margin as well. I mean a bit more short term here about the Q1. I was looking at the history, and it's a bit all over the place, the gross margin direction in Q1. And it seems to be that in an up cycle, it's up quarter-on-quarter. In the down cycle, it's down quarter-on-quarter. The average is down 70 bps, but it's very diverse. So I was wondering if you could help us on the direction of the gross margin given all the moving parts right now for the Q1, please?

For Q1, we see gross margin -- clearly, we don't give indication of the gross margin of the following quarter. But what I can say is that we see gross margin at least at the current consensus for Q1. Let's say, more than that, I would not say more at this stage.

Okay. And maybe one on the satellites. Could you maybe update on -- you mentioned datacenter, but what about LEO into next year? How is it shaping? I mean you kept your guidance for a long time. It's a broad guidance of above $3 billion. But I was wondering if you could give more color on this year and next year as you get more visibility, please?

I pass the question directly to Remi.

We see a lot of momentum across the board, not only in SpaceX, but other programs that are going to be ramping. I would expect actually next year to be as good, if not better than this year in terms of revenue contribution. As a matter of fact, I do believe that we are right now working hard to go and find more capacity to enable even more revenue. So that's the momentum we're into kind of very consistent with what you're hearing from one of our customer's earnings calls in terms of the momentum they're getting.

Sorry, you mean that revenue contribution as a percentage of revenues or just growing on absolute -- just to clarify this point.

My comment was the latter. Revenue would be a lot larger than this year.

The next question comes from Sebastien Sztabowicz from Kepler Cheuvreux. Sébastien Sztabowicz: On a global basis, what kind of visibility you have for the coming quarters and where the backlog is standing at the end of Q3? And attached to that, on the revenue side, how do you see the seasonality building up for '27? Should we expect a better than usual seasonality?

Sebastien, your line was breaking up. Can you rephrase the question what is the seasonality for Q1, you were asking? Sébastien Sztabowicz: No, seasonality for '27. Do you expect better than usual or normal seasonality moving into '27?

Okay, the seasonality will be different. Why? Because as I explained a few minutes ago, we are on the dynamic where we are basically rebalancing the spread of ST across the various vertical. As an example, okay, at the midpoint of the guidance we provide, the automotive vertical's weight of total revenue ST will be about 35%. And then, okay, industrial will be, let's say, between 22%, 24% CCP, so means the digital infrastructure, datacenter, low Earth orbit satellite and some legacy will be slightly above 20% and personal electronics 22%. For 2027, we expect to see CCP continuing to grow. So you see the balance between the verticals will be significantly different. So of course, it will impact the seasonality. So this is some indication we will give clearly in January. And the first... Sébastien Sztabowicz: The first question was on the backlog level today. And what kind of visibility do you have for the coming quarters based on your backlog right now?

You see, okay, with a book-to-bill that was well above 2, okay, and revenue of $3.5 billion. Now our backlog is well above $20 billion. So it's representing compared Q4 midpoint of the guidance, now about 5 quarters of revenue, which start to be appropriate visibility to drive our operation. We are still far the peak of 2023 and 2022.

The next question comes from Menon Janardan from Jefferies.

I'll just start with a sort of clarification and then I have a question after that, but both on gross margin. You had said that you don't expect Q1 gross margin to be at least in line with current consensus. I'm just looking at it both Visible Alpha and Bloomberg, it's about 38%. So are you suggesting that gross margin could decline slightly quarter-on-quarter, given that you are guiding at 39% in Q4? And then the question also is on gross margin. So your target model -- medium-term target model is 44% to 46%. If you take more than 400 basis points from the fab conversion, that means you're sort of saying before that comes through, it's about 41%-ish. So should we assume that between Q4 this year at 39% and, let's say, Q3 2027, your gross margin will essentially go up by, give or take, around 200 basis points? And then as you go beyond that, you will get the full effect of 400, 450 basis points. And when you gave that target model, you did not have so much of visibility on the AI side where you've said that your margin is above the target model on areas like silicon photonics. So given that, that is so strong, is there a chance that your gross margin in this cycle with the help of that as well as pricing could actually end up above that target model?

So basically, Janardan, you want to know the gross margin in Q1 and for 2027, right? I will let Lorenzo answer but [indiscernible] guide for the gross margin going forward. But Lorenzo, do you want to...

No. At the end, clearly, in the past, and as we said also before, clearly, one of our focus is to increase our profitability and increase the gross margin. One of the key element on the gross margin definitely will be our reshaping plan. As we said, our reshaping plan, let's say, will factor in, in our gross margin towards the end of next year and in 2028, let's say, substantially. And this is one of the important element. The other important element is that the growth that we are in AI and the growth in the AI infrastructure and the growth that we have, let's say, now in the industrial that is very significant, as you have seen, is accretive for our gross margin. And this we do expect continuing, let's say, progressively also in 2027, so bringing a positive impact on the gross margin. As I said about the gross margin of the Q1 quarter, at this stage, for me, I don't want to guide, but what I said is that we will be at least at the level of the current consensus. And what's to say more than that, clearly, let's say, there are other elements that we need to consider. And one of the elements that you have to keep in mind is that we are also in an inflationary, let's say, environment for which there is a cost increase in our input cost. This will, let's say, have some impact. But on the other side, there are other positive that are, as I said, the impact of the improvement in manufacturing efficiency, the mix that will enter and also some continued price increase on our top line.

Understood. Just on the first answer, so you suggested and I understand it that because of the strong AI growth, you could end up above the target model. Is that the way I should understand your answer?

The question is if we can end up above our target model, let's say. The point is that for the time being, we target to be at our target model, let's say. And of course, as I said, there are many components. One could be the mix. Let's say, we'll see. But we will, let's say, update you moving forward, let's say, in January and during the next year.

The next question comes from Gianmarco Bonacina from Banca Akros.

Just a follow-up on your book-to-bill, if you can provide a little bit more color for the different verticals. And also, you talked about seasonality for the year. If you can give also some indication given that you indicated you have good visibility also in terms of the Q1, also with the comment you made on the calendar days, we know that normally Q1 is down about 10% sequentially. So if you can indicate directionally this time, given seasonality and visibility, maybe the performance will be different.

So about the book-to-bill that is significantly above 2 for ST overall. It's above 1 for each of our end markets, product group in region. And specifically, it is well, well above 2 for Communication Equipment and Computer Peripheral and Industrial. When I say well, well above 2, I have to say well above 3. About the seasonality, again, what we can say similarly, Lorenzo commented that we feel comfortable to deliver the gross margin at least at the consensus for Q1. We feel comfortable to deliver at least the revenue consensus for Q1 on the calendar we have indicated.

So despite having only 86 days in Q1.

The next question comes from Didier Scemama from Bank of America.

Two questions. One, can you give us a sense of the early indications of your pricing negotiations with your Tier 1 customers, if that -- you want to share anything about that? And I've got a follow-up.

Let's say that -- what we can say is that in terms of pricing, we confirm that in this context that this higher input cost and the increasing lead time, we are increasing pricing on some selected product. This, of course, is an ongoing process. When we look, let's say, about pricing, assuming that this gap between supply and demand remains having -- and this is what we see today when we look at the evolution of our backlog and when we look at the billing, let's say, the fact that next year, we may have in terms of pricing stability or slightly increase, but this is not out of reach. We think that is something that we may see. Anyway, at this stage, it's a little bit early to comment more in general because as you can understand, the discussion for next year, let's say, have not yet been completed. So we are in discussions with our customers.

Yes. No, very clear. Sorry, I just want to understand better what you're trying to say for Q1 because you just said a number of things which are, in my understanding, contradictory to what you said before. So I just wanted to make clear. The number of calendar days in Q1 implies a decline of about 10.6% sequentially, and then you've got normal seasonality of down 10%. So let's say, call it minus 21. Consensus is like minus 5%, minus 6%. And you just said you think you'll be in line with consensus on revenues adjusted for the calendar number of days. So sorry, I'm not sure what to understand from all of that. If you could help me, that would be great. And related to this, the seasonality in Q1 normally is strongly influenced by your Personal Electronics business and by the price cuts that you give to your Tier 1s. So you've answered the second part. On the Personal Electronics side, it looks like your Q3 was a lot better than expected. Is there a risk of a pull-in? Or is it genuine end demand and therefore, no risk that Q1 is worse than seasonal, let's say?

No. I will try to be much more clear, okay. With the visibility we have with a quarter of 86 days, we are comfortable to deliver the consensus of Q1.

Very clear. And on the Personal Electronics, I mean, Q3 strength, was that a pull-in? Or do you think there was just end demand being better than expected? I mean you guided very cautiously for Q3 in fairness.

No, no, no. There is no specific pull-in. It is -- we just deliver, okay, the demand of the customer. And no, we don't see a specific pull-in.

The next question comes from Jakob Bluestone from BNP Paribas.

I just had a question on your silicon photonics ramp. Just wondering if you can give a little bit of an update on how that is progressing in terms of yields? And maybe if you can also update us a little bit on your expansion of your production capacity there.

Thank you. I pass the question to Remi.

Jakob, we -- it's going as per plan. The electro-optical yield our customers are getting at this stage are setting a new bar for the industry. Keep in mind that we are probably the only at-scale 12-inch SiPho player shipping right now. And I think it speaks volume for the value of 12-inch in critical dimension and yield entitlement. So I can say with high confidence that our customers are very, very, very happy with the yield that our technology platform is providing. Clearly, this year, the contribution is not marginal, but it's not at all to the tune of what it will be next year. There will be -- we expect a steep growth, which itself is backed up by the fact that we've communicated that we would increase capacity by 4x year-on-year. Actually, it will be a little bit more than 4x when you look specifically at the number. So I think silicon photonics in ST is going to be a long growth story for which we happen to have actually long-term visibility.

Do you have a follow-up, Jakob?

A quick one, maybe I don't know if you can comment at this point, but just in terms of the -- you obviously had lower OpEx this quarter, obviously flagging a little bit of a step-up for Q4. Just sort of if you can maybe just give us an update on the outlook for your OpEx beyond here? Do you think it could be sort of flattish next year? Or given some of the cost inflationary pressures you flagged, could it be up next year?

On the OpEx of next year, directionally, I would say that they will go up. At this stage, it's a little bit too early to comment. But keep in mind that during 2027, we will keep on continuing our cost savings programs, which should bring, let's say, something in the range of $100 million, $120 million benefit -- we should have also, especially in the second part of the year, some decline in our start-up cost. This means that are impacting our other income and expenses in a negative way. Conversely, on the other side, we are accelerating some of our investment in new business opportunities, as you can imagine. And then on the other side, there will be also, let's say, some impact related to the normal wage salary inflation and these kind of things. So there will be some increase next year, well below, let's say, with anyway an improvement in our expense-to-sales ratio.

The next question comes from Stephane Houri from ODDO BHF.

First question is a clarification on what has been said about the satellite opportunity for next year? Because basically, you're saying you're not sure you will grow, but you're working hard to do so. Can you maybe elaborate and separate a little bit what is the end demand at the moment? And how many customers you have besides your main customer? And what is also linked to the capacity ramp issues?

I'll repeat what I said, Stephane. I said that actually, I expect next year revenue to be same or better than this year revenue. That's the first thing I said. This year revenue was a record year for that business due to the fast growth of our major customers. Next year, that growth continues unabated and comes with additional customers piling on top. And we are actually working hard to release pocket of capacity to be able to capture the entire demand they have because I can say with confidence that at this stage, when we look at front-end module technology, and by BiCMOS technology, there is nothing coming close to it. And keep in mind that actually the technology matter in that case because the noise figures that you can actually get out of the technology will translate into the efficiency of the spectrum, which itself translate into the efficiency of the deployment. So we truly have a first row seat position in those accounts. And right now, I will state that next year, we are most likely going to be capacity gated not demand-gated. And we are working to go and as usual, to find ways to go and release more and more capacity over time. That's the summary of what I said.

Okay. And the second question, the follow-up is about automotive. We didn't speak a lot about automotive today. And there is a sense that the book-to-bill seems to be above 1, so improving. At the same time, the situation among the automotive customers doesn't seem to improve. So what is your view on what's going to happen in 2027? And if you can maybe elaborate a bit about the electromobility opportunity and the software-defined vehicle opportunity at the same time?

Automotive, first of all, the visibility we have from our European Tier 1 is really stable since more than 6 months now, which is really positive compared to the previous year. In '26, we expect to grow for the full year '26 automotive, that is slightly above 10% for us that is basically consistent with the market. It is pretty well spread with our main product group except, of course, radio frequency and optical cable. We are not yet putting optical cable in cars. But between AM&S and Power Discrete and let's say, the digital processing solution, the growth will be pretty similar. But then yes, we are seeing an acceleration of our Power Discrete in Q4 [indiscernible] will be well above 20% -- 25%. So this is driven by the European market of electromobility. So we expect next year the automotive to contribute to the growth of the company. Most likely [indiscernible] I repeat, next year the growth driver of the company will remain the AI datacenter will remain the infrastructure industrial will remain the B2B industrial business for factory automation robotics, this kind of stuff. Our consumer overall, we believe will remain soft or weak. But good confidence level to continue to grow in automotive, driven by European electromobility and also, the many socket wins we have in internal combustion engine cars and hybrid.

Okay. And just one follow-up, a view on silicon carbide for 2027.

Silicon carbide, first of all, this year, okay, we will grow about 20% which is a positive news. So clearly, the main growth driver is European, let's say, electrical car, and this will continue next year. Importantly, for us on silicon carbide, and also part of our, let's say, turnaround of the profitability of this group is the diversification we have engaged first of all, for the power supply of datacenter, but also for power application requesting high power conversion in the industrial market. So next year will be another year of growth, of significant growth for our silicon carbide. And okay, we expect to have the benefits of the wafer size conversion to 8-inch, both from Catania and Chongqing.

We have time for one more question.

The last question comes from Sandeep Deshpande from JPMorgan.

Just quickly coming back to the gross margin question. You've got this 400 bps of gross margin improvement once the reshaping program is over. How much of that is coming from -- today, you're running both the 200- and 300-millimeter fabs. And so what percentage of your gross margin uplift is from that versus how much of it, could you remind us is coming from -- because you're going to have lower cost on the wafer, et cetera, per chip associated with the shift to 300 millimeters? And my second question is that despite your revenue growth, which was quite considerable in both the Analog, MEMS and Sensors business and in your Power & Discrete business, your margin actually declined year-on-year in both those business units. Why was that?

Okay. In terms of, let's say, clearly improvement of above 400 basis points is a combination of different elements. Clearly, the size of the wafer, the fact that we move from fabs that are poorly automatized so with a significant cost of labor on, let's say, fabs that are much more automatized. Now to split, let's say, the various contributor is a little bit complex to be done. When we look, let's say, the profitability of AMS, let's say, overall going down on a year-over-year basis clearly here is the group where is mainly major impact where there is the major impact of the legacy fab that we are now, let's say, on which we are now transferring the products. Clearly, here, there is an impact related to the fact that in this moment, there are some extra cost, as I was explaining overall for the company, but there is also, let's say, an impact on the efficiency of this fab on which we are transferring the cost. So here, when we look at this segment, clearly here is now suffering about this transition phase. It's also true that will be one of the groups that will, let's say, enjoy more when we will finally move to the new fab, to the most advanced fab here, especially for the 300-millimeter. And at the same time, the Power & Discrete, where Power & Discrete will enjoy the move from the 150 to the 200 millimeter.

Just a quick question on this. When you talk about the guidance of 39% gross margin in the fourth quarter, you've highlighted how much of that is -- how much underloading is involved. But how much of this 39% gross margin has impact from this negative impact from this double counting of costs, et cetera?

So for what concern, let's say, the underloading, yes, it's 40 basis points. The extra cost, let's say, related to the new mask, transfer of the process, trials and so on, we estimate in the range of 30 basis points. It's more difficult to say, let's say, the overall inefficiency created by this activity in the fab. It's something that we don't quote because it's a little bit more difficult to really say what is this impact. But this is an impact that, of course, will disappear when we will move, let's say, to the new fab. Anyway, it's 40 basis points, let's say, the unused and 30 basis points, the extra cost. This is what we report usually each quarter.

Thank you, Sandeep, and thank you, everyone. I think that concludes our call. So thank you, everyone, for joining us, and feel free to reach out if you have further questions, the IR team.

Ladies and gentlemen, the conference is now over. Thank you for choosing Chorus Call, and thank you for participating in the conference. You may now disconnect your lines. Goodbye.